Re: Length limit for variable names in your language?

[Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)]

In article <9do3c2$2e$1@news.ox.ac.uk>, "Edward Rosten" <look@my.sig> writes:
> In article <3AFF40AF.C7FB3D9C@uol.com.br>, "Cesar Rabak"
> <csrabak@uol.com.br> wrote:
> 
>> Edward Rosten wrote:
>>> 
>>> > What this limit is for C++,Fortran,Lisp,Pascal etcetcetc ?
>>> 
>>> According to K&R 2nd edition, C guarntees that at least the first 31
>>> characters of an internal name are significant. The standard guarntees
>>> 6 monocase characters for external identifiers, minimum.
>>> 
>>> -Ed
>> 
>> The present standard (a.k.a. C99) changes this to:
> 
> I was referring to the C89(?) standard. C89 compliant (or nearly)
> compilers are still much more common. However things depending on
> external linkage are really at the mercy of the linker and library
> software opn the host system.

In order to meet the Ada length requirements, the Compaq Ada compiler
uses another mechanism that the linker for connections between Ada
modules, using the normal linker mechanism to implement that other
mechanism.  Thus there is no degradation of name length between Ada
modules.

I presume other Ada compilers do something similar, but I am less
familiar with the linking mechanisms in those cases.  Supporting
the full length of identifiers with all characters significant is
a guarantee that is important to Ada programmers, as is portability
between various implementations.